Exploring the Decline of Narrow Networks on the ACA Marketplaces in 2017
Daniel Polsky, Janet Weiner, Yuehan Zhang
SUMMARY: The prevalence of narrow provider networks on the ACA Marketplace is trending down. In 2017, 21% of plans had narrow networks, down from 25% in 2016. The largest single factor was that 70% of plans from National carriers exited the market and these plans had narrower networks than returning plans. Exits account for more than half of the decline in the prevalence of narrow networks, with the rest attributed to broadening networks among stable plans, particularly among Blues carriers. The narrow network strategy is expanding among traditional Medicaid carriers and remains steady among provider-based carriers and regional/local carriers.
The narrowness of provider networks in ACA health plans has drawn attention both as a strategy for reducing premiums and as a potential barrier to access to primary and specialty services. In previous briefs, we have described the breadth of physician networks on the ACA marketplaces in 2014, 2016, and 2017. We found that the overall prevalence of narrow networks for Silver plans in 2017 is 21%, which is a decline from 2016 (25%) and 2014 (31%).
The prevalence of narrow networks could decrease over time for many reasons: existing plans could broaden their networks, plans with disproportionately narrow networks could exit the marketplace, and new plans with broader networks could enter the marketplace. In this brief, we explore these trends and describe differences by state, type of plan, and type of carrier, paying particular attention to the influence of plan exits from, and entrants to, the marketplace.
WHAT WE DID
Network size is estimated for the markets in a state where plans using that network are sold and represents the ratio of the number of physicians participating in a network in those markets divided by the total number of physicians in those markets. As in previous years, we categorized network size into five groups using arbitrary cutoffs that might provide meaningful information to consumers: x-small (< 10%), small (10%-25%), medium (25%-40%), large (40%-60%), and x-large (≥ 60%). We define “narrow” networks as including fewer than 25% of eligible physicians (x-small and small combined).
Because some networks are only attached to a single plan while others are attached to multiple plans in the marketplace, we use the plan as the unit of analysis. To adjust for the fact that some plans are only offered regionally within a state while others are sold statewide, we summarize plans by weighting by the fraction of the state’s population living in counties where the plan was offered. We chose this approach as it reflects consumers’ experiences in choosing between different plans, rather than networks. Because this approach is somewhat arbitrary, we also test whether our findings change substantially using the network as the unit of analysis.
From the 2017 list of all 4,353 qualified health plans (and 72,103 unique plan/county combinations) sold in the marketplaces for all 50 states and DC as provided by the RWJF HIX Compare dataset, we identified 428 unique provider networks offered by 217 different carriers. We used comparable data from 2016 in our analyses. Further details on the development of the analytic dataset are available in our previous briefs. The 2016 data had 5,022 plans and 531 networks. Our data reflect the well-reported ACA marketplace contraction that occurred in 2017. Plan types have shifted as well: we found a 36% decrease in preferred provider organizations (PPOs), from 1,385 to 885; a 4% decrease in HMOs, from 2,643 to 2,536; a 21% decrease in point-of-service (POS) plans, from 528 to 417; and a 10% increase in exclusive provider organizations (EPOs), from 466 to 515. Given this shifting landscape, we focus our analyses on the effects of “churn” on changes in the prevalence of narrow networks.
We compared network size among plans whose carriers participated on the marketplaces in both 2016 and 2017 (“stable” plans) with those that left the marketplace in 2016 or entered the market in 2017 (“churn” plans). By our definition, a plan is considered stable as long as its carrier participated on the marketplaces in both years. While the stable plans are the same plans in both years, the churn plans consist of exiting plans in 2016 and entering plans in 2017.
Note that we defined a unique carrier as a group of issuers offering qualified health plans, which shared a parent company or group affiliation within a given state. We categorized carriers into types based on a set of decision rules, as described in Table 1.
We explored differences overall, by type of carrier, and by type of state marketplace among Silver plans. State marketplace type was based on types reported by the Kaiser Family Foundation. In the analysis, State-based Marketplace includes twelve states with state-based Marketplaces (AR, CA, CO, CT, DC, ID, MA, MD, MN, NM, NV, NY, OR, RI, VT, WA) and four states that use federal platform, but largely manage their own marketplace (AR, NM, NV, OR). Two states—Hawaii and Kentucky—switched state marketplace types from 2016 to 2017, and they were analyzed separately.
WHAT WE FOUND
Figure 1 shows network size in 2016 to 2017, among all plans and among Silver plans. The prevalence of narrow networks declined from 26% to 21%. There were small declines in the fraction of plans with x-small, small, and medium size networks. Among Silver plans, the prevalence of narrow networks and the trends are very similar, as the same networks are typically tied to plans of different metal levels. We restrict the remainder of our analyses to Silver plans.
We examined the degree to which these overall trends are tied to growing network size among stable plans or whether the trend was tied to a different composition of plans and carriers in 2017 as a result of the substantial exit of carriers in 2017. We identified 188 stable carriers that existed in both 2016 and 2017. These stable carriers offered 1,408 Silver plans in 2016 and 1,552 Silver plans in 2017. The plans of the 76 carriers in 2016 that exited in 2017 and the 29 carriers that entered in 2017 make up the exiting and entering plans in the churn category. Exiting plans make up a greater fraction of 2016 plans than entering plans among 2017 plans. Twenty-one percent of Silver plans available in 2016 were not available in 2017 (378 plans) and the 141 entering Silver plans represent only 8% of all Silver plans available in 2017.
Figure 2 shows the changes in network size among Silver plans whose carriers participated in both 2016 and 2017, compared to those in 2016 that exited the market and those that entered the market in 2017. Among the Silver plans from the stable carriers, the fraction of plans with narrow networks declined 2 percentage points — from 21% to 19%. This is a smaller decline than the overall trend among Silver plans where the fraction of plans with narrow networks declined 4 percentage points — from 25% to 21%. This suggests that half of the decline in the fraction of narrow network plans can be attributed to the churn. While the plans from both exiting and entering carriers are relatively narrow when compared to the networks of plans from stable carriers, it is the much higher exit rate among these narrow network plans that explains much of the overall decline in the prevalence of narrow network plans. Among 2016 Silver exiting plans, 24% had x-small networks and 11% had small networks. Among 2016 stable plans, only 6% were x-small and 15% were small. This overall difference in narrowness (35% vs. 21%) is concentrated among x-small networks.
Figure 3 looks at trends by carrier type. The largest changes from 2016 to 2017 were among national carriers, which decreased as a proportion of all plans (Silver plans: 18% to 10%), and among Blues plans, which increased as a proportion of all plans (Silver plans: 35% to 45%). We find substantial decreases in narrow networks among national carriers and Blues plans, and increasing prevalence among regional/local plans and traditional Medicaid carriers.
Figure 4 looks at network size for stable and churn plans for each carrier type. Blues plans were quite stable, with 95% of 2016 Blues plans still offered in 2017. Because of that, the change in network size among Blues plans is due to their decreasing use of narrow networks among stable plans, from 18% in 2016 to 11% in 2017. In contrast, national plans were quite unstable; carriers of 70% national plans exited the marketplace, many taking their x-small networks with them. In addition, when National plans remained in the market, they used narrow networks less frequently. Traditionally Medicaid carriers were stable, with 93% of 2016 plans still offered in 2017. These plans continued and expanded their use of narrow networks, (among stable plans, from 26% in 2016 to 46% in 2017). Provider-based carriers and the regional/ local carriers had mostly stable plans and these plans did not see much of a change in their use of narrow networks. The results within the churn category are highly variable due to small sample sizes in most cases. Of note, however, is that while exiting Medicaid carriers had a high prevalence of narrow networks (83%), there were no narrow network plans among the few entering Medicaid carriers.
Marketplace Type and State-level Trends
In our previous brief, we documented differences in the prevalence of narrow network plans in 2017 by the type of marketplace (as defined by the Kaiser Family Foundation), with narrow networks concentrated in the 12 state-based marketplaces. In Figure 5, we show that state-based marketplaces have an increased prevalence of narrow networks (from 31% in 2016 to 35% in 2017), while states using the federally-facilitated marketplaces have a decreased prevalence of narrow networks (20% to 11%), as have state-federal partnership states (24% to 16%).
Previously, we provided state-specific data on network breadth in 2017 in an online appendix. To examine state-specific trends, we plotted average network size for Silver plans in 2016 and 2017 in Figure 6, with a parity line repressing a state with same average network size in 2016 and 2017. It reveals some variability by state, with growing network size in some outlier states (e.g. KY, VA, OK); however, average network size in 2017 in most states was well within 20 percentage points of their 2016 networks.
WHAT IT MEANS
Carrier exits from the marketplace account for more than half of the decline in the prevalence of narrow networks from 2016 to 2017. The rest can be attributed to broadening networks among stable plans, particularly among Blues carriers. The largest single factor in the decline was that 70% of plans from National carriers exited the market and these plans had narrower networks than returning plans. The narrow network strategy is expanding among traditional Medicaid carriers and remains steady among provider-based carriers and regional/local carriers.
As a sensitivity check, we ran our analyses at the network level, rather than at the plan level, to check for differences in trends. We found a few stronger downward trends, but the results were qualitatively very similar.
We considered whether the use of tiered physician provider networks might change our findings, because we only capture the largest tier in our data. If there is a move towards more tiered physician networks, this trend would not be captured in our analysis. However, a quick analysis of plan cost sharing suggests that physician tiering is rare and shrinking. We found only 6% of plans had a tiered cost sharing plan for generalists and specialists in 2016. This decreased to 3% in 2017.
In sum, our results suggest that carriers with traditionally broad networks (national and Blues) are trending away from their use of narrow networks in this market; in contrast, carriers with experience in narrow networks (such as Medicaid, provider-based carriers, and some local/regional ones) are maintaining their use of narrow networks in their plan offerings. The data suggest signs that the carriers with a greater commitment to narrow network strategies in this market are those with more experience with these networks and perhaps those with stronger connections to the local markets they serve.
Daniel Polsky, PhD, Executive Director, Leonard Davis Institute of Health Economics
Janet Weiner, PhD, MPH, Associate Director for Health Policy, Leonard Davis Institute of Health Economics
Yuehan Zhang, ScM, Statistical Analyst, Leonard Davis Institute of Health Economics
Since 1967, the Leonard Davis Institute of Health Economics (LDI) has been the leading university institute dedicated to data-driven, policy-focused research that improves our nation’s health and health care. Originally founded to bridge the gap between scholars in business (Wharton) and medicine at the University of Pennsylvania, LDI now connects all of Penn’s schools and the Children’s Hospital of Philadelphia through its more than 250 Senior Fellows.